SUMMARY Acute respiratory distress syndrome (ARDS) is associated with high morbidity and at least 25% mortality in both children and adults. Pharmacologic treatment has proven ineffective in decreasing mortality/morbidity in ARDS which is likely due, in part, to heterogeneity within patients presenting with this syndrome. Plasma biomarkers are indicators of underlying pathobiology in ARDS. Latent class analyses of adults enrolled in ARDS network trials and the HARP-2 trial indicate that at least two ARDS subphenotypes exist with differing clinical outcomes and response to treatment. Importantly, biomarkers can distinguish these subphenotypes. The multi-site study, Genetic Variation and Biomarkers in Children with Acute Lung Injury (BALI; R01HL095410), prospectively enrolled patients from the Randomized Evaluation of Sedation Titration for Respiratory Failure clinical trial (RESTORE; U01 HL086622) of children with acute respiratory failure. Of the 549 patients enrolled in BALI, 69% (378) met criteria for pediatric ARDS (PARDS). The goal of the current proposal is to use the BALI cohort to identify subphenotypes in children with PARDS using plasma biomarkers, clinical, and demographic data, and once identified to determine a minimal set of biomarkers and/or clinical data that can define these subphenotypes. This approach was used successfully in studies of adults with ARDS. This proposal will leverage those studies to inform the analysis of children with PARDS. The central hypothesis is that there are subphenotypes in children with PARDS which differ in their underlying pathophysiology. The rationale for this study is that if biomarkers can be used to separate the heterogeneous group of children with PARDS into subphenotypes which differ in pathophysiology, then they may also differ in responsiveness to therapies, as suggested by studies in adults. This proposal addresses one of the missions of the NICHD Trauma and Critical Illness branch which is preventing, treating and reducing critical illness, with respiratory failure being a topic of particular interest. This proposal's Specific Aims are: 1. To identify subphenotypes in children with PARDS using a combination of clinical, demographic and plasma biomarker data from the BALI and RESTORE studies. 2. To determine whether relevant clinical outcomes differ between subphenotypes and whether response to the targeted sedation strategy tested in the clinical trial differs between subphenotypes. This proposal will provide proof of concept that children with PARDS can be stratified into groups with differing risk and/or potential responsiveness to targeted therapies and will inform the next large, prospective ancillary studies examining whether PARDS subphenotypes exist and differ in response to therapy. The identification of PARDS subphenotypes could identify PARDS patients at highest risk for morbidity or death, and would have a major impact on future trial design, potentially allowing more precisely targeted treatments.